Real-time journalism: Implications for news writing

Real-time journalism: Implications for news writing

Karla Aronson

Wall Street Journal articles are longer than realtime articles on the Dow Jones Business Wire, but readability levels are nearly the same.

The financial news industry has delivered the news directly to users by telegraph since the mid-1860s and more recently by computers since the 1960s, eliminating the mass medium as intermediary. Today, financial news wires transmit news instantaneously (in real-time) to subscribers around the world via high-speed computer and telecommunication networks.

Geared primarily to business coverage, these real-time services operate in all time zones 24 hours a day. More than 500,000 users – primarily in financial institutions, business corporations and government agencies – subscribe, paying $100 on average for the news. Some 13 news agencies, including Reuters PLC, Dow Jones & Co., Knight-Ridder Inc. / and Bloomberg L.P., comprise the real-time media industry with estimated revenues of $500 million.l

Despite the forecast that the electronic delivery of news and information will be routine in the 21st century,2 however, researchers know little about this highly profitable news industry – partly because the failure of early experiments3 in electronic journalism resulted in few subjects for study and partly because real-time ventures may have gone unnoticed because of their limited distribution. How does the content vary from newspapers? Is it as journalistically sound and as readable? The influences of real-time production processes and media routines on content are examined here and characteristics of articles in real-time and in newspapers are compared.

Results may shed light on an industry that, while owned by the world’s largest news agencies, primarily serves an elite subscriber base. As scholars call attention to “the pressures that are transforming the goals, tools, professional standards and sociopolitical function of the major journalistic media,”4 realtime financial news industry represents one such transformation that may become the model for future, more general, computerized news services. Studying the product could tell editors and publishers much about their own products as well as what continued movement to specialized, niche-based products might have in store as far as changing not only the product but the process needed to produce it.

Theory and background

Most news media content faces five levels of influence: individual, routines, organizational, extramedia and ideological.5 Simply put, the individual level involves the professional background and personal characteristics of journalists. Meanwhile, three factors determine media routines: the audience as news consumers, the media organization as news.

producers and sources as suppliers of news. The structural and functional relationships of media employees determine media organization while the extramedia level refers to outside influences, such as sources, special interest groups, PR campaigns, advertisers, readers and government, among others. Finally, the ideological level evaluates whose interests media routines and organization serve and the limits those interests place on news perspectives.

The five levels form a hierarchy of influences that can and do occur simultaneously. Each one’s relative importance varies, but the model permits investigation into decisive influences. This study focuses on the influence of media routines – “those patterned, routinized, repeated practices and forms that media workers use to do their jobs.”6 – to examine how audiences, sources and the media organization affect real-time news production.

Audience needs

Audience-oriented media routines address what is acceptable to the audience. News criteria help predict what an audience will find appealing and important. For example, news judgment often is based on news values which commonly include: prominence/importance, human interest, conflict/controversy, the unusual, timeliness and proximity.7 By comparison, real-time news delivery serves audience needs for timely information, considered important to the financial markets. Competition in real-time news delivery is measured in fractions of a second; a minute behind with the news is officially late. As one realtime news editor said, “Timeliness is everything. There was a time I thought AP (Associated Press) was a fast service, but they catered to the daily newspaper market. Competition is by seconds. Ten years ago it was by hours.”8 As a result, real-time media routines oriented toward timeliness and important financial market information would be expected to influence the amount and rate of news delivery to the audience.

A second audience-oriented media routine concerns news presentation techniques and formats that standardize story arrangements and headlines. Real-time news stories are transmitted on a computer screen, limiting story lengths to about 15 to 20 lines of text – or about two or three short newspaper paragraphs of information – per screen. Real-time news articles are generally released with a headline followed by five to seven short paragraphs of information, about two computer screen-pages long – limits that would constrain a narrative story structure. But real-time stories can be released in several newstakes as separate news items, permitting a story to be developed at greater length. Editing for the computer screen involves more than just tight writing to conserve computer screen space. Editors must arrange the facts in a straightforward, logical order, selecting the most important information and omitting the rest.9

In addition, the real-time medium may intensify the need to condense news for the computer screen and, thus, for its audience. Reporters and editors have only enough time and space to convey the essential information, and that is usually all the reader wants to see. Since real-time news is directed toward a sophisticated audience of financial professionals, background material and explanation often aren’t needed. With speed as the primary aim, writing style becomes secondary. But another audience need – readability – remains crucial. Audience ability to immediately comprehend the material is of utmost concern to real-time journalists. “If the trader doesn’t understand you, first he’s going to curse you. Then he has to figure out what it means and that takes time,” said one such reporter.lo Ease of reading material is determined by several elements, including typography, reader interest and writing style,l’ of which the latter is a concern to this study.

Readability measures often evaluate vocabulary and grammatical complexities,lz two items also emphasized by research. For example, an early study concluded that the more difficult concepts conveyed in hard news resulted in longer words and longer sentences in the news stories.l3 And another study indicated a gradual decline in news readability over a century, mainly because of journalistic use of longer words.

Meanwhile, the newspaper’s role has changed with the advent of broadcast media, providing less narrative description of events and more difficult expository interpretation and explanation.4 Providing explanation and background material may solely be a print media function. For instance, a content analysis of videotex and newspaper articles found that the former were more likely to deal with immediate or breaking By comparison, realtime news content also provides more of a news bulletin service, possibly leaving the role of providing explanation and background material to the newspaper.

Still, newspaper articles must be written to many reader levels. The financial news articles in the Wall Street Journal, for instance, serve different purposes for more than two million readers. In contrast, real-time media are written for a highly educated, specialized subscriber base that makes practical, daily use of the news. The contention that media routines address audience needs might indicate greater readability levels for real-time news articles because of the audience’s need for quickly and immediately comprehensible information. Real-time readability also may be easier, given the limited computer screen space and the continuous dissemination of shorter-length articles. But real-time media’s primary emphasis on speed necessitates quicker writing and editing and could lead to decreased readability levels when compared to newspaper articles.

Organizational needs

As complex organizations with regular deadlines, the news media need systems to cope with the unpredictable and infinite number of occurrences they encounter and thus must routinize work to control Media routines help organizations manage time and space constraints. For example, news production deadlines and space or time available to present the news affect the amount and length of content.l7

In the real-time industry, the central sites and times of news are linked across space and time through global computer networks, which expand the spatial and temporal scope of newswork by linking newsrooms and financial markets around the world and across different time zones. Reporting and editing schedules, as well as news deadlines, must be coordinated on an international basis. A real-time executive described this coordination and expansion: “The real-time medium demands relevance, speed and accuracy because of the time element. Trading activity is minute-to-minute. Real-time journalists are writing as it happens, and within seconds you see the (financial) markets react.”18


As news suppliers, sources influence content by dictating routines via news conferences, news releases and public relations campaigns. Sources also can restrict access and regulate information release.19Similarly, financial reporting requires expert knowledge of financial markets and institutions. Financial journalists develop their own knowledge of market activity but must extensively rely on expert sources to provide and interpret news.

Although news media theoretically have countless resources available, they heavily depend on interviews. Newsgathering conventions also tend to favor routine channels of news sources; staff and time limits encourage journalists to actively pursue only a small number of sources available and suitable in the past.20 In financial news coverage, real-time media may rely upon the same routine sources of news, yet may encounter different source constraints.

For example, a minute-tominute news environment requires a different reporting style. Real-time journalists depend on insights of sources actively engaged in financial market activity and can be in frequent contact with these sources throughout the day. Such heightened time pressure limits the time available to gather news from multiple sources. To illustrate, consider the study that found print media used more sources because of broadcast’s more limited story lengths and tighter deadline constraints. Each medium favored particular source types, resulting in different story angles.2′

In summary, the constraints of media routines should evidence themselves in real-time news content. This study will examine the characteristics of real-time financial news for that purpose and answer the following research questions:

1. How do newspaper and real-time financial news articles compare

2. How do newspaper and real-time financial news articles compare

2. How do newspaper and real-time financial news articles compare in source ability?

3. How do newspaper and real-time financial news articles compare in source attribution?


The authors selected the content of Dow Jones & Company’s financial newswires and the Wall Street Journal to characterize the real-time medium and to compare its news articles to the newspaper articles. As the largest supplier of real-time financial news in the United States, the company distributes its real time news to about 190,000 computer terminals and publishes the country’s largest newspaper. Despite the Dow Jones focus, findings may be representative of other real-time agencies because all serve the same customers, reporting essentially the same events by staffs that see each other at the same events and who trade notes about techniques; staffers commonly have worked at two or more different agencies.22Final considerations in the selection of Dow Jones news content were access to newswire articles and the electronic text articles of the Journal through the Dow Jones News/Retrieval (DJNR) online database.23

The sampling was conducted in two stages. First, the authors retrieved from the Dow Jones Business News Wires an approximate hour of news story text that covered the closing of the New York Stock Exchange, which was also expected to be an active news period.24 Second, this study was limited to market coverage and company news to ensure a reliable comparison of strictly financial reporting, as opposed to spot news, enterprise stories and government reports. Dow Jones Business News Wires on May 10,1994 and the following day’s issue of the Journal were arbitrarily selected for comparison.

All company stories published in the newspaper (excluding an insert) were selected to match with the real-time news articles. Of the 60 stories in the newspaper, 49 were matched with real- time articles – amounts that, for the descriptive purposes of this study, the authors deemed appropriate. A larger sample size would have been more desirable, but data availability precluded any control of that aspect by the authors. Market comment articles, published in the third section of the Journal, were selected to compare with the real-time market comments. Of the six such stories in the newspaper, five were compared with real-time articles.

For the first research question, the samples were analyzed for length in terms of the number of words, sentences and paragraphs per article. To answer the second research question, the Flesch Reading Ease and Flesch-Kincaid Grade Level tests were applied.25

For the final research question, the authors counted the number of attributed news sources identified either by direct quotation or by citation in the samples. The number of attributed sources identified by name was compared to the number of unidentified, or anonymous, sources in each medium. Intercoder reliability was 1.00 for the identified sources category and .84 for the anonymous sources category.26



In the DJNR system, 18 headlines are displayed on a single screen (screen-page). Each headline is numbered anywhere from 1 to 999, with the newest item always assigned No. 1. The user accesses the story by typing its headline number. The headlines also provide the source of the news, its release time and a textual summary of the subject. Real-time stories are written in takes or versions, and usually limited to about 45 lines (1.5 screens) of text. As mentioned above, DJNR distributes news in a cycle of 1 to 999 items. The sample cycle was distributed in about 3.5 hours, or at a rate of 6.8 cycles per 24 hours, or nearly 7,000 news items. The average number of items was about 285 per hour, or about 4.7 per minute.


The first research question sought to compare how newspaper and realtime financial news articles in length; in this study, the average number of words, sentences and paragraphs for 54 newspaper and 120 real-time articles was calculated.

As Table 1 shows, the newspaper articles contained an average of 297 words, compared with 181 for real-time stories. The average number of newspaper sentences was slightly more than 14 per article, compared with 8.5 for real-time stories . The average number of newspaper paragraphs was 7.5, compared with 5.5 for real-time articles. The ratio of words, sentences and paragraphs of the newspaper articles to the real-time articles was 1.64,1.68 and 1.37, respectively, nearly half-again more. All the differences were statistically significant.27

The authors also calculated the lengths of the company stories and the market comment news articles in each medium. As the table shows, the newspaper company stories averaged 237 words, 11.2 sentences and 5.9 paragraphs, compared with 168 words, 7.8 sentences and 5.1 paragraphs for the realtime company articles. The newspaper-to-real-time company story ratios for words and sentences and paragraphs were 1.41 and 1.44, respectively. The average number of paragraphs were not statistically different.

For the market comment newspaper articles, the stories averaged 889 words, 44 sentences and 23.2 paragraphs, compared with the real-time market comment averages of 226 words, 10.8 sentences and 6.5 paragraphs. The newspaper-to-real-time market comment ratios for words, sentences and paragraphs were 3.94, 4.09 and 3.55, respectively. All the differences were statistically significant.


The second research question sought to determine how newspaper and real-time financial news articles compare in readability. This study used the Flesch tests measuring two levels of reading difficulty: semantic level of vocabulary – which counts the average number of characters per word – and syntactic level of sentence construction, which measures the average number of words per sentence.

In addition, reading ease was computed. The scores range from 0 to 100. Standard writing averages 60 to 70; the higher the score, the greater the number of people who readily can understand the text. The grade level score indicates a grade school level (e.g., 8.0 mean an eighth grader would understand the test); standard writing approximately equates to seventh-to-eighth-grade levels.

As Table 2 indicates, the average reading ease scores were not significantly different in newspaper (47.2) and real-time (48.9) articles. There was no difference in the company stories category, as both media recorded reading ease scores of about 45. And the market comments story category also showed only a slight difference. The similarities continued for grade level scores in each category, with the only differences between the two media coming in company stories; real-time company stories had a slightly higher grade level score. Semantic and syntactic levels also showed no significant variation, although the latter seemed to provide the most contrast. As the table shows, semantic levels were similar across the two media for market comment stories ( 5 .12 for newspaper and 5.07 for real-time) and only slightly different (5.35 and 5.29, respectively) for company stories.

Meanwhile, syntactic level differences were broader but still not significantly different; i.e., overall, the newspaper measured 21.4 words per sentence, compared with 21.9 for real-time articles. Among company stories, the averages were 21.4 for newspapers and 22.1 for real-time, while market comments showed averages of 20.7 and 21.1, respectively.


The third research question sought to compare newspaper and realtime financial news articles in source attribution.

Overall, Table 3 shows the 54 newspaper articles averaged 1.4 news sources, compared with 0.9 sources for the 120 real-time articles -1.6 times as many but not a significant difference. Much of this was because of the company stories, where newspaper articles averaged 1.4 times as many news sources as did real time stories, although this again was not a significant difference. The outstanding differences came in the market comment stories, where the newspaper averaged 5.4 attributed sources, compared with 1.4 for the real-time medium, a statistically significant difference.

This trend continued when the numbers were categorized by identified and unidentified sources. For example, overall the newspaper articles averaged 0.8 identified sources compared with 0.5 for the real-time stories; that meant the newspaper averaged 0.6 unidentified sources, compared with 0.3 for the realtime medium; but again, these differences were not significant. Again, the company story category reflected this difference most closely (0.6 identified sources for the newspaper, 0.5 for real-time; 0.4 and 0.2 unidentified, respectively).

Still, the only significant differences were found among market comment stories, where newspaper stories averaged 2.2 identified and 3.2 unidentified sources compared with 0.6 and 0.8, respectively, for real-time stories. The difference disappeared, however, when article lengths were held constant. Interestingly, only in the market comment stories did either medium use more unidentified than identified sources.


This research focused – and successfully utilized a communication model28 – on a single level of media routines’ influences on news content. Other levels obviously need testing in real-time media; e.g., journalists’ backgrounds and perceptions, organizational structure, economic structure and ideological concerns. This study’s findings tend to lend support to contentions29 that routines affect content; several practical explanations account for the results.

As to length, real-time news articles were shorter on average than the newspaper articles in number of words, sentences and paragraphs. Real-time editorial policy imposes maximum story lengths for easier reading on the computer screen. Shorter article lengths satisfy real-time audience needs for brief, factual information that can be read quickly. Time constraints to produce news also prohibit longer stories. Brief items are released as separate news takes over time. Computer software and hardware impose their own limits on space, although each real-time provider would have such limits. In contrast, newspaper articles consolidate information into a single story for daily publication. Such shorter lengths also might account for the greater amount and faster rate of real-time news delivery.

Regarding the nearly equal readability measures, real-time articles despite shorter length – do not appear to be more or less difficult to read than do newspaper stories. But both media appear more difficult than standard reading. This may indicate the complexity of financial subject matter.30 Still, neither realtime media’s greater time pressure nor real-time’s essentials-only content was enough to generate readability level differences.

Journalists – real-time or otherwise – still have to write simply. And both media use the same format conventions. Story organization and narrative structure have been found to be shared among all media31 and endure in the realtime medium. Moreover, some of the stories in both media appeared to originate from the same sources. In some instances, when bylines were included, the same author was identified in the real-time and newspaper articles. Also, it seems much of the content originated from wire sources because two compared articles were so often similar. It seems both media may simply edit the material for their own formats.

Finally, practical considerations probably influence the usage of sources. Just as broadcast news reports use fewer sources than newspaper stories,32 realtime – with its similar time-sensitive orientation toward continuous production of shorter news items – also uses fewer sources. This practice is considered acceptable largely due to the sharper time and smaller space constraints. And when story lengths are considered, the differences in source usage disappear. A broader base of sources would be preferred, but it is not demanded. The need to identify sources does not appear as vital to establishing real-time story credibility; it is generally customary for sources to provide information off the record. And just as newswires used fewer sources than newspapers because of technologies, deadline constraints and work routines,33 the real-time medium’s sources may be less readily available to respond within the more immediate real-time deadlines, when compared with newspapers’ longer deadline cycle.

In addition, news gathering promotes reliance upon routine channels of news sources. The time limits of news production encourage journalists to actively pursue only a small number of sources who are available and adequate.34 Similarly, real-time journalists – working under heightened deadline pressures – may be less dependent on a broad range of news sources.

When it comes to attributing sources, real-time newspapers may be less concerned with identifying them. One real-time reporter, who also wrote a daily column for the newspaper, admitted as much.35 Also, as such reporters acquire greater expertise in financial news, they may be less reliant on news sources.

Overall, these findings have significant implications for journalism practices. As for editors and publishers of the current traditional print product, the indicated differences between print and the real-time product suggests they should re-examine their desires concerning current story lengths and how they might translate in the real-time medium. Will shorter be better or desirable? Can current reporters be taught to write in shorter drafts?

In addition, the real-time medium reorganizes journalistic routines. As a computer-oriented medium, it allows a great quantity of articles to be released, many simultaneously. Such continuous release and global, ongoing demand for news makes for shorter, quicker stories, which in turn may necessitate fewer sources or – at the least – reliance on regular, more available sources in some cases. When contemplating a possible addition of real-time editions, can editors plan and coordinate such changes in routine and adequately prepare their staffs?

Additional research needs to determine whether this translates into more one-sided, less objective stories. As a result, the entire concept of objectivity might need re-examination within the evolving framework of electronic media, e.g., do standard notions of objectivity apply? This in turn might suggest a broader concept of newsworthiness for electronic journalists or, at least, a redefinition of the term.

Moreover, the rapid and continuous release of shorter stories in realtime raises additional questions about the nature of the information communicated, i.e., do real-time practices detract from the quality of content? Some might argue that smaller-sized computer screens make the shorter-story finding expected. But the fact that real-time stories often are developed in multiple takes/releases leads to interesting questions about how real-time stories are edited in contrast to traditional printed stories. Qualitative analysis is needed to determine whether real-time coverage ever surpasses superficiality and approaches the depth of its print counterpart.

From the consumer’s standpoint, this study doesn’t address the utility of such coverage. Similarly, retention and comprehension need further exploration, in light of the readability levels detected. Though the similarity of those levels in the two media would imply that journalists – in whatever medium don’t tailor their prose for different audiences, surprisingly real-time editors stress that the newspaper requires added explanatory material and avoidance of jargon that real-time does not.36So the two similar readability levels might suggest closer scrutiny of real-time writing practices and surveying the realtime reader as to the effect of this similarity.

In summary, this study supports the notion that the real-time medium alters journalism practices. Journalism news gathering and writing practices are affected by the speed and computer format of real-time news delivery. Realtime source identification is similar but stories are shorter. These differences perhaps point to the increasing influence of technology on the routines of news writing – the building of stories in several takes and their immediate dissemination contrast with methodically integrating information into a single story and arranging that information in the most critical order. A fast food analogy comes closest to describing the two methods, with real-time being fast and traditional practices representing home cooking. The inevitable question arises as to which food is better – for journalism and its readers?


1. Robert Murphy, Financial Newswires: The Real Time News Agencies 1992-1993. New York: Waters Information Services,1992, p. 8.

2. M.L. Stein, The Future of Electronic News. Editor & Publisher, May 2,1992, pp. 7071. See also James Carey and John Pavlik, Videotex: The Sword in the Stone in Demystifying Media Technology. California: Mayfield Publishing Co.,1993, pp. 163168.

3. David Weaver, Videotex Journalism. New Jersey: Lawrence Erlbaum Associates, 1983.

4. Jay Blumler, ed. Symposium: Journalism in Crisis and Change. Journal of Communication, March 1992, pp. 3-107.

5. Pamela Shoemaker and Stephen Reese, Mediating the Message: Theories of Influences on Mass Media Content. New York: Longman, 1991, pp. 3-4. 6. Shoemaker and Reese, op. cit., p. 85. 7. Shoemaker and Reese, op. cit., p. 91.

8. Sam Alcorn, deputy managing editor. Dow Jones Capital Markets Report, Dow Jones & Co., New York, March 29, 1993. Personal interview. 9. Hilary Goodall and Susan Smith Reilly, Writing for the Computer Screen. New York: Praeger Publishers,1988, p. 53.

10. Pomesh Sahi, reporter. Dow Jones & Co., New York, March 29, 1993. Personal interview.

11. Robert Hoskins, A Readability Study of AP and UPI Wire Copy. Journalism Quarterly, Summer 1973, pp. 360-363.

12. Researchers measure vocabulary complexity by determining the proportion of long or common words in the text; texts with a high share of long words generally are more difficult. Measuring grammatical complexity involves determining average length of sentences; more difficult texts generally have longer sentences. 13. Wayne Danielson and Sam Dunn Bryan, Readability of Wire Stories in Eiqht News Categories. Journalism Quarterly, Winter 1964, pp. 105-106. 14. Wayne Danielson, Dominic Lasorsa and Dae Im, Journalists and Novelists: A Study of Diverging Styles. Journalism Quarterly, Summer 1992, pp. 436-446. 15. Natalie A. Brown and Tony Atwater, Videotex News: A Content Analysis of Three Videotex Services and Their Companion Newspapers. Journalism Quarterly, Autumn 1986, p. 558.

16. Shoemaker and Reese, op. cit., p. 97.

17. Gaye Tuchman, Making News: A Study in the Construction of Reality. New York, Free Press, 1978, p. 41.

18. Bruce Chalfant, assistant vice president of information services, Dow Jones & Co., New York, June 7,1994. Personal interview. 19. Shoemaker and Reese, op. cit., p. 150.

20. Shoemaker and Reese, op. cit., p. 106. See also Leon V. Sigal, Sources Make the News, in Reading the News. New York: Pantheon Books,1986, p.17 and Gans, op. cit., p.116.

21. Dominic Lasorsa and Stephen Reese, News Source Use in the Crash of 1987: A Study of Four National Media. Journalism Quarterly, Spring 1990, pp. 60-71. See also Frederick Fico, Replication of Findinqs in Newspaper and Wire Service Source Use in Statehouse Coverage, Newspaper Research Journal, Fall 1985, pp. 45-51. 22. Murphy, op. cit., p. 131.

23. The five proprietary Dow Jones Business Newswires are Dow Jones News Service (real-time news on U.S. and Canadian companies, industries, U.S. government agencies, financial markets, the stock market and the economy); Dow Jones International News Services (global real-time business, financial and economic news); Professional Investor Report (news, trading alerts and statistical reports on active stocks, plus coverage of overall market activity); Dow Jones Capital Markets Report (real-time coverage of all major sectors of the international fixed-income and financial futures markets); and Federal Filings (real-time notification of filings with the Securities and Exchange Commission and detailed analysis).

24. The New York Stock Exchange – instead of NASDAQ – was chosen because NYSE operates with a trading floor in a defined time period. NASDAQ is a 24-hour computerized stock quote/trading system that has no physical trading floor. The chosen NYSE trading period signifies an active trading time in the U.S. market day during which the real-time news was generated.

25. See Ron Smith, How Consistently Do Readability Tests Measure the Difficulty of News Writing? Newspaper Research Journal, Summer 1984, pp.1-8. The Flesch test measures two levels of reading difficulty, the semantic level of vocabulary and the

syntactic level of sentence construction. The former counts the average number of characters per word; the latter measures the average number of words per sentence. Reading ease scores range from 0 to 100 (standard writing averages about 60-70); the higher the score, the greater the number of people who can readily understand the text. 26. Oleg Holsti, Content Analysis for the Social Sciences and Humanities. Reading, Massachusetts: 1969. No set level of intercoder reliability was established because of the exploratory nature of this study. See Klaus Krippendorf, Content Analysis: An Introduction to Its Methodology. Beverly Hills, California: 1980, pp.146-148. 27. The authors used an independent t-test, which uses the averages of independent groups – in this study, newspaper and real-time formats – to determine statistical significance (i.e., whether the difference in the group averages are due to chance or random error).

28. Shoemaker and Reese, op. cit. 29. Shoemaker and Reese, op. cit. 30. See Danielson and Bryan, op. cit., pp.105-106. 31. Herbert Gans, Deciding What’s News: A Study of CBS Evening News, NBC Nightly News, Newsweek and Time. New York: Pantheon Books, 1979. 32. Lasorsa and Reese, op. cit. 33. For example, Fico, op. cit. 34. For example, Sigal, op. cit., and Gans, op. cit. 35. Tom Lauricella, reporter, Dow Jones & Co., New York, July 25, 1994. Personal interview.

36. Sam Alcorn, deputy managing editor. Dow Jones Capital Markets Report, Dow Jones & Co., New York, June 7,1994. Personal interview.

Aronson was a Master’s degree student in the department of journalism at the University of Texas-Austin at the time this article was written. She now works for Thomson Financial Networks. Sylvie is assistant professor and Todd professor at the University of Texas.

Copyright E.W. Scripps School of Journalism Summer/Fall 1996

Provided by ProQuest Information and Learning Company. All rights Reserved